now that we have a routine to clean up the packet-specific infos in the
usb conversation, there's no need to clean up usb_conv_info after parsing the configuration descriptor this wouldn't work anyway if one of the subfunctions throws an exception Change-Id: Ifd13f320db1ec17c3b706778ce0f5dffa1349770 Reviewed-on: https://code.wireshark.org/review/3729 Reviewed-by: Martin Kaiser <wireshark@kaiser.cx> Tested-by: Martin Kaiser <wireshark@kaiser.cx>
This commit is contained in:
parent
d6775dc40e
commit
5dc742b35f
1 changed files with 0 additions and 12 deletions
|
@ -2004,18 +2004,6 @@ dissect_usb_configuration_descriptor(packet_info *pinfo _U_, proto_tree *parent_
|
|||
|
||||
proto_item_set_len(item, offset-old_offset);
|
||||
|
||||
/* Clear any class association from the Control endpoint.
|
||||
* We need the association temporarily, to establish
|
||||
* context for class-specific descriptor dissectors,
|
||||
* but the association must not persist beyond this function.
|
||||
* If it did, all traffic on the Control endpoint would be labeled
|
||||
* as belonging to the class of the last INTERFACE descriptor,
|
||||
* which would be especially inappropriate for composite devices.
|
||||
*/
|
||||
usb_conv_info->interfaceClass = IF_CLASS_UNKNOWN;
|
||||
usb_conv_info->interfaceSubclass = IF_SUBCLASS_UNKNOWN;
|
||||
usb_conv_info->interfaceProtocol = IF_PROTOCOL_UNKNOWN;
|
||||
|
||||
return offset;
|
||||
}
|
||||
|
||||
|
|
Loading…
Reference in a new issue